Multiple Bernoulli relevance models for image and video annotation

By Shaolei Feng, Raghavan Manmatha, Victor Lavrenko. Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004., 2004

935 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/05e410576620107762a616e51c1738fa1a0953f6

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

ImageCLEF 2014: Overview and Analysis of the Results (2014, methodology): “Particularly, in the training step the conditional probability of an image, being labeled by every class, is calculated using some parametric [40],[6] or nonparametric [34] methods.”
Retrieval-Enhanced Machine Learning (2022, background): “It has shown successful results in a wide range of retrieval tasks [2, 8, 14, 28, 36, 68, 73], demonstrating the quality of the produced query representations for retrieval.”
Automatic Caption Generation for News Images (2013, methodology): “The literature is littered with various attempts to learn the associations between image features and words using supervised classification [1], [2], instantiations of the noisy-channel model [3], latent variable models [4], [5], [6], and models inspired by information retrieval [7], [8].”
A Discriminative Kernel-Based Approach to Rank Images from Text Queries (2008, methodology): “A longer survey could also have described alternative models such as the Multimodal Hierarchical Aspect Model [4], [3], the Multiple Bernoulli Relevance Model [16], or Latent Dirichlet Allocation [5].”
XRCE ’ s participation to ImagEval (2007, methodology): “In contrast to this model where the annotation words for any given image are assumed to follow a multi- nomial distribution , Feng et al [13] proposed to model them with a multiple Bernoulli model and the image feature probabilities using a non-parametric kernel density estimate.”
Semantic Annotation and Retrieval of Music and Sound Effects (2008, background): “A comparative summary of alternative supervised one-versus-all [11] and unsupervised ([12], [13]) models for image annotation is presented in [1].”
Visual Information in Semantic Representation (2010, methodology): “For example, a common ﬁrst step to all automatic image annotation methods is partitioning the image into regions, using either an image segmentation algorithm (such as normalized cuts; Shi and Malik 2000) or a ﬁxed-grid layout (Feng et al., 2004).”
Supervised Learning of Semantic Classes for Image Annotation and Retrieval (2007, methodology): “The various methods differ in the definition of the states of the hidden variable: Some associate a state to each image in the database [13], [21], while others associate them with image clusters [3], [12], and some model higher-level groupings, e.”
On the Role of Correlation and Abstraction in Cross-Modal Multimedia Retrieval (2014, background): “For example, in [29] a query text, and in [30] a query image is used to retrieve similar text documents and images, based on low-level text (e.g., words) and image (e.g., DCTs) representations, respectively.”
Topic Models for Image Annotation and Text Illustration (2010, methodology): “, 2002), and methods inspired by information retrieval (Feng et al., 2004; Lavrenko et al., 2003).”
A Survey on Visual Content-Based Video Indexing and Retrieval (2011, methodology): “[8] use the multiple-Bernoulli distribution to model image and video annotation.”
Baselines for Image Annotation (2010, methodology): “The top portion of the table displays published results from several top-performing methods that approach the annotation problem from different perspectives, using different image representations: CRM (Lavrenko et al. 2004), InfNet (Metzler and Manmatha 2005), NPDE (Yavlinsky et al. 2005), MBRM (Fen”
Joint Video and Text Parsing for Understanding Events and Answering Queries (2013, methodology): “In contrast to the bag-of-words representation widely used in the video understanding and multimedia literature (e.g., [1], [2], [3], [4]), in this work we propose to represent the joint interpretation of video and text in the form of a parse graph.”
Topic regression multi-modal Latent Dirichlet Allocation for image annotation (2010, background): “Several modeling variations have been proposed, see [1] [2] [9] [8], with the different forms of probability distribution assumed for caption words (multinomial vs. bernoulli) and image features (mixture of Gaussian vs. non-parameteric density estimation).”
Im2Text and Text2Im: Associating Images and Texts for Cross-Modal Retrieval (2014, background): “This limitation has led to a growing interest in the problem of automatic image annotation [4, 6, 7, 10, 19, 35, 36, 38].”
Automatic tag expansion using visual similarity for photo sharing websites (2010, background): “Recent studies in automatic annotation of visual data is promising [1–3, 5– 7, 9, 12, 14, 20, 23, 24, 26, 35, 36].”
How Many Words Is a Picture Worth? Automatic Caption Generation for News Images (2010, background): “…(Vailaya et al., 2001; Smeulders et al., 2000), instantiations of the noisy-channel model (Duygulu et al., 2002), latent variable models (Blei and Jordan, 2003; Barnard et al., 2002; Wang et al., 2009), and models inspired by information retrieval (Lavrenko et al., 2003; Feng et al., 2004).”
Translating topics to words for image annotation (2007, unknown): “In the second category of approaches [5, 4, 8, 9].”